Yıldız Tolga
Department of Psychology, Istanbul University, Laleli, Istanbul, 34134, Turkey.
Integr Psychol Behav Sci. 2025 Jan 2;59(1):10. doi: 10.1007/s12124-024-09876-2.
This theoretical paper offers an in-depth examination of the intersection between Theory of Mind (ToM) and artificial intelligence (AI), drawing on developmental psychology and philosophical analysis. By investigating the key developmental stages at which children begin to understand that others have distinct mental states, the paper provides a framework for assessing the cognitive boundaries of AI systems. It critically interrogates the pervasive human inclination to anthropomorphize machines, particularly through the attribution of complex mental states like "knowing," "thinking," or "believing" to AI entities that lack subjective experience. The paper argues that AI, while capable of simulating cognitive processes, operates without the conscious awareness that defines human cognition, raising profound epistemological and ethical questions. It explores the broader implications of this projection for society, considering how our conceptualization of AI affects both technological development and social structures. Ultimately, this interdisciplinary inquiry calls for a more nuanced understanding of the distinctions between human and machine cognition, advocating for responsible approaches to AI as its capabilities evolve.
这篇理论性论文利用发展心理学和哲学分析，对心理理论（ToM）与人工智能（AI）的交叉领域进行了深入研究。通过调查儿童开始理解他人具有不同心理状态的关键发展阶段，本文提供了一个评估人工智能系统认知边界的框架。它批判性地质问了人类普遍存在的将机器拟人化的倾向，特别是通过将诸如“知道”“思考”或“相信”等复杂心理状态归因于缺乏主观体验的人工智能实体。论文认为，人工智能虽然能够模拟认知过程，但它的运行缺乏定义人类认知的有意识的觉知，这引发了深刻的认识论和伦理问题。它探讨了这种投射对社会的更广泛影响，考虑我们对人工智能的概念化如何影响技术发展和社会结构。最终，这种跨学科探究呼吁对人类和机器认知之间的区别有更细致入微的理解，随着人工智能能力的发展，倡导对其采取负责任的方法。
